When Ethereum Gas skyrockets to 100 gwei, users not only face transfer cost pressures but also suffer from high interaction fees with some mainstream DEXs and lending protocols. The real culprit behind this is often overlooked—the traditional oracle push model.

In this mode, as soon as prices fluctuate slightly, dozens of nodes simultaneously push data onto the chain, even if no transactions occur. This is undoubtedly a huge waste of block space.

So is there a better solution? The pull model offers an alternative approach:

First, data is stored off-chain. Only when a transaction is initiated is the necessary price data "pulled" onto the chain. This prevents redundant data accumulation and network congestion. Second, in terms of response speed, the pull mode directly accesses contract logic from node memory, avoiding the inherent latency of the push mode. This is especially critical for high-frequency traders—because that brief latency window is an opportunity for sandwich (MEV).

But how to solve the practical problem that lending protocols require real-time price monitoring? A hybrid solution has emerged—retaining push for the lending sector to ensure security, while switching high-frequency derivatives trading to the pull mode. This maintains the stability of the lending system while reducing costs for traders.

Next time you encounter a project still using traditional oracles, consider asking: with the optimization options in front of us, why not choose them?